During the past grant cycle, working in Araihazar, Bangladesh, where there is a wide range of both arsenic (As) and manganese (Mn) in drinking water, we have demonstrated that both elements have adverse effects on child intelligence. The current proposal seeks to further test and elaborate on the hypotheses that exposure to As and Mn are associated with deficits in intellectual function in two distinct population samples. The first draws on a U.S. population, for whom we are able to apply standardized scores to estimate the magnitude of the association with IQ. The second draws on a Bangladesh population to examine whether or not the effects of As on intellectual function are reversible;it takes advantage of a Columbia Earth Institute initiative to provide As-free deep tube-well water to an exposed population in Araihazar, Bangladesh. In New Hampshire (NH), we will conduct a cross-sectional study of 500 fourth-grade children to examine the relations between water As, water Mn and IQ. We selected schools to yield a sample in which roughly 70% will consume water from wells with <10 ug As/L, 30% from wells with >10 ug/L, based on USGS data. The distribution of Mn in wells is less well characterized, but limited USGS data indicates that roughly 10% of wells in this region will have Mn concentrations above the EPA Health Advisory Level of 300 ug/L. In Bangladesh, we will test the hypothesis that provision of low As/low Mn tube wells is associated with improved intellectual function. We propose a two-year prospective study of 300 children (baseline ages 7-9) in Araihazar, Bangladesh. For each of the following 4 groups, 75 children will be recruited: a) low As, low Mn;b) low As, high Mn;c) high As, low Mn;d) high As, high Mn. Based on EPA guidelines, the definition of "low As" will be <10ug/L, and "low Mn" will be <300 ug/L After their initial intellectual assessment, children with high As wells will be provided access to low As/low Mn community wells and education to promote their use. Children will be assessed before remediation and at 12- and 24-months post-remediation. Given that in the U.S., approximately 1.3 million people are exposed to >10 ug/L As in drinking water, and household Mn, though less well characterized, is known to be elevated in several states (e.g. PA, NH), identification of adverse effects on childhood intelligence may have significant impact both on public health and on establishment of policies to reduce exposure.